And therefore the difference in tweeter edge difference will be smaller than with a straight edge.
If the straight edge would be very close to the tweeter, yes, that is having a similar effect of spreading diffraction issues, but it should also be taken into account under which angle the partial edge diffractions are most noticeable and leading to cancellation effects.
this seems a special form of sound then, as typical waves "see" all kinds of edges, especially the sharp and big ones.
If they are radiated in more of less unattenuated form into the direction of the obstacle with the latter being big enough to cause a diffraction, surely. My point was that the bridge is causing a significant level of diffraction for the waves going to the sides, in parallel to the main baffle, already, so they are presumably not going to see these edges being further away with sufficient level to cause annoying edge diffraction.
And yes a surround or a basket ring of a woofer is an obstacle for treble waves, posing a certain risk of causing diffractions. But the situation with conventional coaxial designs using the midrange cone as a waveguide, is a bit more sensitive to such diffractions as the waves are originating from the inner of the cone, basically following the shape of the midrange cone. That is not the case here.
a super small baffle (but the Geithains are neither)
The bridge is in fact a very small baffle as seen from the tweeter, except for its width which explains the edge diffraction in a narrow band from the sides.
Geithain speaks about "cardioid" up to 300 Hz and no mentioning of the mid being cardioid n their page. So it its not even clear what is "intended".
With ´intended´ I meant directivity control to achieve a more or less constant directivity index without a significant decrease in the band the midrange has kicked in (500..1.2K Hz).
My guess would be they are relating to specific ´cardioid´ for those models which were available with cardioid or omnidirectional bass in different versions. But I cannot speak for the manufacturer as I am not affiliated in any way nor do I have any sympathy for all their design choices. Found the cardioids with smaller woofers (particularly 8") as well as the very big midwoofers in 2-way designs and their tweeter arrays rather questionable.
What is in your opinion the cause of such "FR issues" under specific angles? If it is not a result of diffraction, what else is there?
Can be anything, reflections from midrange/tweeter hitting the woofer diaphragm, interference effects between drivers, cone breakup for the midrange, imperfections regarding midrange cardioid behavior.
Wide directivity gives better spaciousness, tonal richness and as Sean Olive puts it, wide directivity speakers disappear better in rooms.
Not sure about that. If wide horizontal dispersion patterns comes with increased level of discrete reflections from the side walls, wider directivity is more likely to result in ´diffuse spaciousness´, tonal imbalance, compromising localization stability and increasing the risk of the speakers themselves being accidentally localizable.
To make wide directivity speakers disappear better, a certain amount of attenuation or diffusion of such reflections is necessary which is more likely to be found in a studio control room rather than a living room.
If your 8361s are projecting a ´compact´ staging, I would prescribe this rather to a lack of diffuse energy >2K in the room causing the phantom sources to appear detached from the ambience/reverb. With a balanced in-room reverb, they can sound as wide as a wide dispersion speaker. Speaking of Genelec, I found the 8341 to be the ideal compromise between these two opposing concepts.